CEOCODILIANS, LIZARDS, AND SNAKES. 557 



may lie oitlier compressed longitudin.illy, or perpi-udiciilarly to the jaws {Dicrodon, 

 Teius). Pterygoid tcetli are but seldom present, and if so, but feebly developed. 



In the most highly develojied forma the skull presents the typical Lacertilian or 

 L'ionocrauian structure; but as we approach the Chakidine forms the skull becomes 

 more depressed, the vacuity between the ante and jjostorbital ]»ortion,s and conse- 

 quently the interorbital septum become much reduced, the arches weaker, and at 

 last we have in Ophiognomon a skull where the pterygoids are largely in contact 

 with the sphenoid, the arches are very weak, the columella is so reduced as to bo 

 hardly distinguishable, and the whole skull forms an almost continuous solid mass. 



The premaxillary boue is single, the nasals double, the frontal and parietal sin- 

 gle; a small supraorbital bone exists in some Ameivas; the palatines are in contact 

 anteriorly; the maxillarj'is excluded from the inl'raorbital fossa, which is bounded 

 by the palatine, pterygoid, and transverse bone, as in Varanus. The headshields are 

 always free from the cranial ossification, a character which separates sharply this 

 family from the Old- World Lacertidte. 



Limbs or their rudiments are present; the degradation begins with i'rof<oj>oroM.y, 

 which has very short limbs, but with five well-developed clawed digits; the limbs 

 are still more reduced in size in Scolecosaurus, and the digital formula is incomplete, 

 although claws are still present; in Cophias and OpMoanomon digits are either 

 entirely lost or reduced to bud-like rudiments; and in Propiif: the pelvic limbs have 

 altogether disappeared. The clavicle is dilated and perforated proxinially ; how- 

 ever. Cope mentions Tretioscinvus and Scolecosaurus as forming exceptions, having a- 

 simple clavicle; the interclavicle is cruciform. 



I have already noted a peculiarity in the visceral anatomy of this 

 family, in the absence of tlie urinary bladder. Tbis structure is 

 alleged by sucb authors as Stannius to be present in all Sauria. I 

 have Ibund no trace of it in the genera Tupiiiamhis, Braccvna, Ameiva, 

 CneinidophoruK, Centropyx, Gymnoplitlialmus, and Callopistes. In all 

 these genera the kidneys have a peculiar form, being expanded trans- 

 versely for the anterior half of their length, and then tapering poste- 

 riorly to a point. In this respect they differ from any family of lizards 

 known to me. 



The alimentary canal presents the usual three parts, stomach, small 

 intestine, and rectum, with terminal rectal cloaca. There is no especially 

 distinguished colon except in Callopistes. In JJracccna a portion of the 

 canal immediately succeeding the stomach is constricted off from the 

 small intestine for a moderate distance. In Callopistes there is a short 

 wide colon without cjiecum. The liver is not elongate, and has only 

 one posterior notch. Recurved lobe of the right lobe inconspicuous or 

 wanting. The ventral mesentery extends beyond tlie liver to a part 

 of the small intestine. 



The mesenteric attachments of the liver are very characteristic in the 

 Teiida'. There is but one suspensor, a median gastrohepatic, but tbis 

 bifurcates above the middle of the organ and each half diverges, and 

 adhering to the caudal margin, extends to the lateral inferior body wall, 

 on each side. In Tupinamhis these sheets are united on the median 

 line for a distance posterior to the liver. The lungs are each attached 

 to the stomach by a'^eparate sheet. The left hepato-parietal sheet is 

 always present in tbis family, but the right one is feeble in some genera 

 and is easily ruptured, as, for instance, in Cnemidophorus. 1 have 



